Today for AI
HOT RADAR
llmHEAT 15.8°

AI CURATED HOT STORY · 10/30/2026

Edge0 Releases Open-Weight Audio8-TTS-Realistic Model

1 reports archived1 independent sourcesupdated 10/30/2026, 13:12:58
Synthesis & Latest Updates
1 Sources Cross-Validated

Edge0 released Audio8-TTS-Realistic, a 0.6B open-weight TTS model under Apache 2.0. It employs a learned realism critic to optimize rhythm and pauses, achieving a 4.86% average error rate on CV3 across 9 languages, outperforming several competitors despite weaker performance in low-resource languages.

LATEST/Edge0 released Audio8-TTS-Realistic, a 0.6B open-weight TTS model under Apache 2.0. It employs a learned realism critic to optimize rhythm and pauses, achieving a 4.86% average error rate on CV3 across 9 languages, outperforming several competitors despite weaker performance in low-resource languages.

Hourly heat curve

5 fully observed hours
Now
15.8
Peak
17.114:00
24h change
–

The line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.

Coverage timeline

Total 1 reports · Latest first
  1. Marktechpost (@Marktechpost)T2·78 pts

    Edge0 Releases Open-Weight Audio8-TTS-Realistic Model

    • The 0.6B parameter model supports 23 languages and 44.1 kHz audio output with fully open weights and code (Apache 2.0).
    • Key innovation involves training a 'realism critic' optimized via GRPO to capture human-like features like rhythm and pauses that WER misses.
    • Achieves a 4.86% average error rate on CV3, beating VoxCPM2 and CosyVoice3, though low-resource languages like Thai remain weak at 48.57% WER.